Systems and methods for biometric identity and authentication
US10878072B2 · kind B2 · utility
Assignee
Inventors
Key dates
| Filing date | Nov 19, 2018 |
| Grant date | Dec 29, 2020 |
| Priority date | — |
| Expiry date | Mar 2, 2039 |
Classification
- Technology area (CPC H)Electricity
- CPC primaryH04W12/06
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
In accordance with some embodiments, a training method for biometric identity and authentication is provided. The method includes obtaining biometric data from a plurality of sources. The method further includes extracting a plurality of feature vectors from the biometric data. The method also includes determining a plurality of identifiability scores correspondingly associated with the plurality of feature vectors, where each of the plurality of identifiability scores provides a quantitative characterization of a relative uniqueness of a corresponding one of the plurality of feature vectors. The method additional includes determining run-time authentication neural network parameters based on a function of the plurality of feature vectors, where the run-time authentication neural network parameters enable extraction of one or more feature vectors from biometric data of a particular user, and the run-time authentication neural network parameters are associated with the plurality of feature vectors determined to satisfy an error threshold.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.